config files missed for training naive text-to-image model #2185
Description
Describe the bug
I used the following command to train a naive text-to-image model and the training process goes well.
Here are files in the model folder
pytorch_model.bin optimizer.bin scaler.pt
random_states_0.pkl scheduler.bin
However, during inference, it raises errors on missing config files.
OSError: Error no file named model_index.json found in directory path_to/text2Euro-naive/checkpoint-4000.
I manually download model_index.json from https://huggingface.co/stabilityai/stable-diffusion-2-1/blob/main/model_index.json and put it into the model folder.
A new error occurred:
Error no file named scheduler_config.json found in directory path_to/text2Euro-naive/checkpoint-4000.
How can I find these missed files for model inference?
Reproduction
export MODEL_NAME="stabilityai/stable-diffusion-2-1-base"
export TRAIN_DIR="./data/train"
export OUTPUT_DIR="./text2Euro-naive"
accelerate launch --mixed_precision="fp16" train_naive_T2I.py \
--pretrained_model_name_or_path=$MODEL_NAME \
--train_data_dir=$TRAIN_DIR \
--output_dir=$OUTPUT_DIR \
--train_batch_size=1 \
--gradient_accumulation_steps=4 \
--gradient_checkpointing \
--max_train_steps=30000 \
--learning_rate=1e-05 \
--max_grad_norm=1 \
--lr_scheduler="constant" --lr_warmup_steps=0 \
Inference
from diffusers import StableDiffusionPipeline
import torch
model_path = "path_to/text2Euro-naive/checkpoint-4000"
pipe = StableDiffusionPipeline.from_pretrained(model_path, torch_dtype=torch.float16)
pipe.to("cuda")Logs
OSError: Error no file named model_index.json found in directory path_to/text2Euro-naive/checkpoint-4000.
> Error no file named scheduler_config.json found in directory path_to/text2Euro-naive/checkpoint-4000.System Info
diffusersversion: 0.10.2- Platform: Linux-5.13.0-39-generic-x86_64-with-glibc2.31
- Python version: 3.9.15
- PyTorch version (GPU?): 1.13.0+cu117 (True)
- Huggingface_hub version: 0.11.1
- Transformers version: 4.25.1
- Using GPU in script?:
- Using distributed or parallel set-up in script?: